Every piece of disruptive technology in history has a life story. For X-BAT, that story begins nearly forty years before its conception with the development of a unique propulsion technology.

AVEN^®, or the Axisymmetric Vectoring Exhaust Nozzle, is the thrust-vectoring engine nozzle that lets Shield AI’s X-BAT launch and land vertically without a runway. It wasn’t initially designed to enable the world’s first AI-piloted vertical takeoff and landing (VTOL) strike fighter. It was created to explore a new avenue of thrust vectoring capability, advancing from 2D to omnidirectional nozzles for enhanced maneuverability. Decades later, that enhanced maneuverability is critical for X-BAT to launch and land anywhere. No runway, no problem.

The Origins of AVEN

AVEN was born out of a simple question that confronted aircraft engineers in the early 1990s: What if the engine itself could increase aircraft maneuverability?

For decades, military aircraft had relied almost exclusively on conventional aerodynamic surfaces — ailerons, flaps, rudders — to maneuver. Those surfaces work exceptionally well throughout much of an aircraft’s flight envelope, but their effectiveness changes with speed and operating conditions. As fighters were pushed to become more agile for demanding air combat missions, engineers began searching for another source of control.

The answer was thrust vectoring.

Rather than relying solely on airflow over wings and control surfaces, thrust vectoring allows an aircraft to redirect the engine’s exhaust, using the engine itself to help maneuver the aircraft. So, engineers turned to designing vectoring nozzles to integrate with existing engines.

Two-dimensional (rectangular) nozzles were created, employing relatively flat flaps to direct the engine’s thrust in a single axis (i.e. pitch or yaw, but not both). While single-axis thrust vectoring does things that traditional nozzles can’t, 2D nozzles had limitations that hindered their effectiveness; they could vector thrust in only a single plane, added immense weight on the engine, created flow losses, or required bulky, complex nozzle designs that added weight, reduced efficiency, and complicated maintenance.

That’s when GE Aerospace devised a different approach – the Axisymmetric Vectoring Exhaust Nozzle.

Instead of converting the engine’s naturally circular exhaust into a two-dimensional (rectangular) nozzle, GE Aerospace’s AVEN preserved the engine’s axisymmetric flow while enabling multidirectional thrust vectoring in any direction. The design retained the efficiency of a conventional variable-area convergent-divergent nozzle while adding the ability to precisely direct thrust. Just as importantly, it was conceived as a system that could be adapted to existing nozzle configurations rather than requiring an entirely new propulsion architecture. It was simple in operation, light in weight, and relatively cheap to manufacture.

The nozzle flew as part of a thrust-vectoring research program on a specialized F-16 as part of the Multi-Axis Thrust Vectoring (MATV) program. It was tested at Edwards Air Force Base, California, in the 1990s, accumulating 87 hours of ground testing and 135 flight hours across 95 sorties. The program proved that an axisymmetric vectoring nozzle could deliver reliable, multidirectional thrust vectoring without sacrificing performance throughout the aircraft’s operating envelope.

The Operational Necessity of AVEN

Just as AVEN was born out of a generation-defining operational problem, so was X-BAT: runways are increasingly vulnerable.

In modern war gaming, more aircraft are lost on the ground than in the air. The defense industry has spent the last 30 years developing increasingly stealthy aircraft to improve survivability in the air, yet those same aircraft remain highly vulnerable on the ground. Fixed infrastructure is easier to target than ever before, and projecting combat power increasingly depends on the ability to disperse aircraft across ships, islands, highways, and expeditionary locations. Yet, legacy aircraft that make up our forces today cannot fill that dire need alone.

VTOL is the fundamental enabler of ground survivability. From the beginning, we knew X-BAT needed to launch and recover vertically while maintaining the speed, range, payload, and survivability expected of a tactical strike platform. VTOL allows us to launch and land almost anywhere, transforming the world into a runway instead of confining operations to a limited number of fixed airbases. By eliminating dependence on runways and enabling constant mobility, VTOL fundamentally improves survivability on the ground. Meeting those requirements, however, made propulsion one of the program’s greatest technical challenges.

During vertical takeoff, hover, and landing, the propulsion system becomes the aircraft’s primary means of maintaining attitude. The demands placed on the nozzle are entirely different from those experienced during conventional forward flight. The propulsion system must be able to execute rapid, precise control motions to operate vertically.

As we evaluated propulsion concepts, AVEN stood out. Its multidirectional thrust vectoring enables the propulsion agility needed for an expeditionary footprint, and its flight history gave us further confidence that the underlying hardware worked. Instead of beginning with an entirely new propulsion system, we could build upon decades of engineering knowledge and focus our efforts on adapting the technology for an entirely different mission.

Giving AVEN a Second Life

The original AVEN program had already answered some of the hardest questions in propulsion development. GE Aerospace engineers had spent years designing, testing, and flying the nozzle, ultimately demonstrating its function through hundreds of hours of testing. That represented an enormous body of engineering work Shield AI didn’t have to recreate.

But the AVEN that flew in the 1990s was designed to improve maneuverability in forward flight. X-BAT asks AVEN to do something fundamentally different. During vertical takeoff and landing, the nozzle becomes the aircraft’s primary flight control system, continuously redirecting thrust to maintain attitude with a level of speed and precision the original program had no need for testing. Converting AVEN to meet those demands meant rethinking how the nozzle actuates, interfaces with the engine, and communicates with the aircraft’s flight controls.

X-BAT’s propulsion system is simple: a single jet engine mounted down the center of the aircraft, similar to an F-16.

At takeoff, the engine ignites its afterburner to generate the thrust needed for vertical launch, while AVEN precisely vectors that thrust for control. Many modern fighter jets already generate more lifting force than their own weight but typically use that extra power for takeoff runs from a runway. X-BAT instead leverages this same capability to lift off vertically, while still achieving performance on par with today’s fighter aircraft. For landing, the challenge shifts to modulating thrust to closely match the aircraft’s weight, with AVEN continuously adjusting thrust direction to maintain a stable, controlled descent.

Working alongside GE Aerospace’s Edison Works team, we took the AVEN direct from a storage warehouse — “Indiana Jones-style,” as we like to say — refurbished its original hardware, integrated it with the F110-GE-129E engine, and returned it to operation. Our teams completed integration, actuation, and engine light-off testing of this propulsion system at GE Aerospace’s Peebles Test Operation, validating that the nozzle, engine, actuators, and control systems functioned as a single integrated propulsion system. It marked the first fully integrated AVEN test campaign since the original flight program more than three decades ago.

Our joint success demonstrated that a propulsion technology developed in the 1990s could be successfully adapted to support a next-generation autonomous aircraft. More importantly, it cleared the path toward X-BAT’s vertical flight testing, where AVEN will transition from an experimental thrust-vectoring nozzle into the propulsion system that enables the aircraft to lift off, hover, transition, and land.

Starting with flight-proven hardware dramatically accelerated X-BAT’s development timeline. Rather than spending years developing and qualifying an entirely new nozzle architecture, we could focus our engineering efforts on solving the challenges unique to vertical flight and modifying the nozzle to the mission set. That allowed us to move from integration to ground testing in a fraction of the time while reducing technical risk along the way.

Beyond AVEN Integration

To me, AVEN is one of the most exciting aspects of the X-BAT program. We aren’t simply preserving an important piece of aerospace history; we’re extending it. By combining decades of propulsion engineering with modern autonomy and a completely new aircraft architecture, we’ve transformed a technology built for one mission into the foundation for something its original designers could never have envisioned.

We may have already successfully integrated AVEN with the F110-GE-129E for X-BAT development, but the work doesn’t end there. The AVEN testing today isn’t the final version. Like every component on X-BAT, it will continue to evolve. The joint SAI/GE Aerospace team is reengineering the nozzle around modern materials, modern guidance and control systems, and modern production techniques, while keeping the original, flight-proven design as its baseline. Future iterations will become faster, lighter, more responsive, and optimized for autonomous vertical flight. As X-BAT matures, so will the propulsion system that powers it.

What’s remarkable isn’t just that AVEN continues to improve. It’s that a technology originally conceived to explore the limits of fighter maneuverability has become the foundation for an entirely different generation of aircraft. Its first chapter proved that multidirectional thrust vectoring could work. Its second chapter is proving what that capability can enable.

About the Author:

Shiva Vallabhaneni is a senior propulsion engineer for X-BAT at Shield AI, where he leads the development and integration of the aircraft’s propulsion system. He brings expertise in aircraft propulsion, thrust vectoring, and system integration. Prior to joining Shield AI, Shiva worked on propulsion testing and development at SpaceX, supporting five Starship launches from engine development through flight. He holds bachelor’s and master’s degrees in Aeronautical and Astronautical Engineering from Purdue University.